An in-vehicle image module is mounted on an automotive vehicle, and detects another vehicle near the vehicle, illumination intensity of the outside of the vehicle and the like. Conventionally, an in-vehicle image module includes multiple image elements as a unit having the same sensitivity, which are arranged to be in a square lattice pattern.
In the conventional module, another vehicle around a vehicle having the module is completely captured by image elements in the module having a sufficiently large number. Specifically, the whole outline of the other vehicle is completely captured. Accordingly, when a light corresponding to a light image of the other vehicle disposed around the vehicle entered into a convex lens in the module from obliquely beneath along with a light path is detected, the light image can be captured with a sufficiently large image range, i.e., sufficiently wide view angle and with sufficiently high resolution. However, a background such as the Sun and a cloud around the vehicle is not completely captured. Further, a space among the Sun, the cloud and the other vehicle is captured by many image elements. Thus, when the light corresponding to illumination intensity around the vehicle entered into the convex lens from obliquely above along with a light path is detected, a sufficiently large image range, i.e., a sufficiently wide view angle is not secured. Further, a region such as the space among the Sun and the cloud, the region which is not required to have high resolution, is captured by many image elements. Accordingly, there is much waste of the image elements. Thus, the view angle is not sufficiently large so that the illumination intensity is hardly detected. Further, since the imager size is small, the dynamic range and the sensitivity are not sufficient so that the illumination intensity is hardly detected. Furthermore, a load for processing the image corresponding to the region not required to have high resolution becomes large, so that a processing time becomes longer.